US7649953B2ExpiredUtilityPatentIndex 91
Differential multiple-length transmit and reception diversity
Est. expiryFeb 13, 2023(expired)· nominal 20-yr term from priority
Inventors:BAUCH GERHARD
H04L 1/0068H04L 1/0618H04L 25/067H04B 7/0669H04L 2001/0096H04B 7/0891H04L 27/34H04L 1/08H04L 1/0071
91
PatentIndex Score
44
Cited by
25
References
22
Claims
Abstract
The present invention achieves differential transmit diversity and related diversity reception schemes transmit symbol constellations which are extended to multiple levels. Heretofore, a group of transmission bits is divided into a first sub-group of transmission bits and a second sub-group of transmission bits. Then, the first sub-group of transmission bits is mapped ( 14 ) onto constellation points of a differential transmit diversity coding scheme from an orthogonal design for coding of the first sub-group of transmission bits. Then, after differential encoding of the constellation points ( 32, 34, 36, 40 ) a length of the transmission symbol vector is scaled ( 38 ) for coding the second sub-group of transmission bits. Therefore, the present invention overcomes restrictions implied through unit length requirements for previously known differential transmit diversity schemes from orthogonal designs.
Claims
exact text as granted — not AI-modified1. A method of achieving differential multiple-length transmit diversity from orthogonal designs using at least two transmit antennas, comprising the steps:
dividing a group of transmission bits into a first sub-group of transmission bits and a second sub-group of transmission bits;
mapping the first sub-group of transmission bits onto constellation points of a differential transmit diversity coding scheme from an orthogonal design for coding the first sub-group of transmission bits;
determining transmission symbols through differential encoding of the constellation points and previously transmitted transmission symbols for setup of a transmission symbol vector; and
scaling a length of the transmission symbol vector for coding the second sub-group of transmission bits;
wherein the step of mapping the first sub-group of transmission bits comprises a step of:
mapping the 2·log 2 (M 1 ) transmission bits onto a constellation vector [A k B k ] of the differential transmit diversity coding scheme according to:
A k =d 2t+1 d (0)*+ d 2t+2 d (0)*
B k =−d 2t+1 d (0)+ d 2t+2 d (0)
wherein
d i are constellation elements of a M 1 -ary phase shift keying PSK modulation scheme;
d(0) is a freely selectable reference point of the M 1 -ary phase shift keying PSK modulation scheme; and
the length of the constellation vector is unit length |A k | 2 +|B k | 2 =1;
the step of determining transmission symbols (S 2t+1 S 2t+2 ) through differential encoding is achieved according to:
( S 2t+1 S 2t+2 )= A k ( X 2t−1 X 2t )+ B k (− X* 2t X* 2t−1 ),
wherein
t is an index in time; and
(X 2t−1 X 2t ) is a transmission symbol vector according to a previously transmitted matrix; and
the step of scaling the transmission symbol vector length is achieved according to:
( x 2t+1 x 2t+2 )=√{square root over ( a q k )}·( s 2t+1 s 2t+2 )
wherein
a is a constant; and
q k ε{−M 2 +1, −M 2 +2, . . . , 0, 1, ., M 2 −1} is a length exponent.
2. A method according to claim 1 , wherein
the group of transmission bits comprises 2·log 2 (M 1 )+log 2 (M 2 ) bits,
wherein
M 1 is the number of possible bit sequences of the first sub-group of transmission bits;
M 2 is the number of possible length values of the transmission symbol vector, and
wherein
the step of dividing the group of transmission bits comprises the steps:
selecting a number of 2·log 2 (M 1 ) bits in the group of transmission bits for the first
sub-group of transmission bits; and
selecting a number of log 2 (M 2 ) bits in the group of transmission bits for the second
sub-group of transmission bits; and
the step of scaling a transmission symbol vector length is based on:
calculating a set of M 2 scaling factors δ k ε{0, . . . , M 2 } from a set of bit patterns [p 1 , . . . , p log2 (M 2 )] covering permutations of the second sub-group of transmission bits for all ρ i ε{0, 1} according to:
δ
k
=
∑
i
=
1
i
=
log
2
(
M
2
)
[
p
i
·
2
i
]
;
and
establishing a pre-determined relation between bit patterns of the second sub-group of transmission bits [u k,2,1 , . . . , u k,2,log2 (M 2) ] and the set of scaling factors, wherein u k,2, =is the i-th bit in the second group of transmission bits.
3. The method according to claim 2 , wherein the step of scaling comprises a step of calculating the length exponent according to:
mapping transmission bits of the second sub-group of transmission bits [u k,2,1 , . . . , u k,2,log2 ,(M 2) ], to a scaling factor δ k according to the pre-determined relation established between bit patterns of the second sub-group of transmission bits [u k,2,1 , . . . , u k,2log2 (M 2) ] and the set of scaling factors; and
calculating the next length exponent according to
q k =δ k −s ( Q k−1 +δ k −M 2 )− M 2 ;
wherein
s( ) is a step function having a value of 1 for non-negative arguments and a value of 0 for negative arguments; and
Q k−1 length exponent representing the absolute length of the transmission symbol vector transmitted prior to calculation of the length exponent q k .
4. A method of differential multiple-length diversity reception of transmission symbols using at least one reception antenna, wherein transmission symbols carry information being coded through mapping of a first sub-group of transmission bits onto constellation points of a differential transmit diversity scheme from an orthogonal design for coding the first sub-group of transmission bits, subsequent differential coding of the constellation points and previously transmitted transmission symbols for setup of a transmission symbol vector, and scaling of a length of the transmission symbol vector for coding a second sub-group of transmission bits, the method comprising the steps of:
organizing transmission symbols into a plurality of reception vectors according to a Pre-determined scheme;
combining the reception vectors for determination of at least a first decision variable and a second decision variable in relation to the first sub-group of transmission bits and further for determination of a third decision variable in relation to the second sub-group of transmission bits;
determining a first detection output in relation to the first sub-group of transmission bits on the basis of the first decision variable and the second decision variable, respectively; and
determining a second detection output in relation to the second sub-group of transmission bits on the basis of the third decision variable; wherein
organizing transmission symbols into a plurality of reception vectors is achieved according to:
y
k
=
[
y
2
t
-
1
(
1
)
y
2
t
(
1
)
*
⋮
y
2
t
-
1
(
n
R
)
y
2
t
(
n
R
)
*
]
,
y
k
+
1
=
[
y
2
t
+
1
(
1
)
y
2
t
+
2
(
1
)
*
⋮
y
2
t
+
1
(
n
R
)
y
2
t
+
2
(
n
R
)
*
]
,
y
_
k
=
[
y
2
t
(
1
)
-
y
2
t
-
1
(
1
)
*
⋮
y
2
t
(
n
R
)
-
y
2
t
-
1
(
n
R
)
]
,
wherein
t is a time index;
n R is the number of reception antennas;
* is a complex conjugate operator; and
y i (j) is a symbol received at time i at reception antenna j.
5. The method according to claim 4 , wherein the step of combining the reception vectors for determination of a first decision variable ŷ 1 , a second decision variable ŷ 2 and a third decision variable ŷ 3 is achieved according to:
y
^
1
=
y
k
H
y
k
+
1
;
y
^
2
=
y
_
k
H
y
k
+
1
;
and
y
^
3
=
y
k
+
1
H
y
k
+
1
y
k
H
y
k
;
wherein
H is operator of transposing a vector and applying the conjugate complex operator * to all vector elements.
6. The method according to claim 4 ,
wherein the step of determining the first detection output in relation to the first sub-group of transmission bits is a hard output detection step; and
the first detection output is determined as constellation vector [A(i)B(i)] from the differential transmit diversity scheme which is closest to a vector set up from the first decision variable and the second decision variable [ŷ 1 ŷ 2 ] according to:
(
A
^
k
B
^
k
)
=
arg
min
i
{
y
^
1
-
A
(
i
)
2
+
y
^
2
-
B
(
i
)
2
}
.
7. The method according to claim 6 , further comprising a step of obtaining the first sub-group of transmission bits through demapping from (Â k {circumflex over (B)} k ).
8. The method according to claim 4 , wherein the step of determining the second detection output in relation to the second sub-group of transmission bits is a hard output detection step; and
the second detection output is determined by a length exponent a qk which is closest to ŷ 3 according to:
â q k =arg min| ŷ 3 −a q(i) |, q ( i ) ∈{− M 2 +1, . . . , −1, 0, 1, M 2 −1}
wherein
{−M 2 +1, . . . , −1, 0, 1, M 2 −1} is a set of all candidate length exponents; and
a is a constant.
9. The method according to claim 4 ,
wherein the step of determining the first detection output in relation to the first sub-group of transmission bits is a soft output detection step; and
log-likelihood ratios for the first sub-group of transmission bits are determined according to
L
(
1
)
(
u
^
k
,
l
)
=
log
p
(
u
k
,
l
=
+
1
❘
y
^
1
,
y
^
2
)
p
(
u
k
,
l
=
-
1
❘
y
^
1
,
y
^
2
)
wherein
k is a time index;
a vector of the first Sub-group of transmission bits u k of dimension 2log 2 (M 1 ) is mapped onto one of the M 1 2 constellation elements of the differential transmit diversity scheme and u k,l is a transmission bit at position l in u k ;
ŷ 1 is the first decision variable;
Ŷ 2 is the second decision variable;
p(u k,l =+1|ŷ 1 , ŷ 2 ) is a conditional probability for u k,l =−1 in view of determined decision variables ŷ i and ŷ 2 ;
p(u k,l =−1|ŷ 1 , Ŷ 2 ) is a conditional probability for u k,l =−1 in view of determined decision variables ŷ i and ŷ 2 ; and
L (1) (u k,l ) is the soft output for the first sub-group of transmission bits.
10. The Method according to claim 4 ,
wherein the step of determining the second detection output in relation to the second sub-group of transmission bits is a soft output detection step; and
log-likelihood ratios for the second sub-group of transmission bits are determined according to
L
(
2
)
(
u
^
k
,
l
)
=
log
p
(
u
k
,
l
=
+
1
❘
y
^
3
)
p
(
u
k
,
l
=
-
1
❘
y
^
3
)
wherein
k is a time index;
u k,l is a transmission bit at position l in a vector u k of dimension log 2 (M 2 ), the vector
u k being set up from the second sub-group of transmission bits;
ŷ 3 is the third decision variable;
p(u k,l =+1|ŷ 3 ) is a conditional probability for u k,l =+1 in view of the decision variable ŷ 3 ;
p(u k,l =−1|ŷ 3 ) is a conditional probability for u k,l =−1 in view of the decision variable ŷ 3 ; and
L (2) (u k,l ) is the soft output for the second sub-group of transmission bits.
11. An apparatus for achieving differential multiple-length transmit diversity using at least two transmit antennas, comprising:
a dividing unit adapted to divide a group of transmission bits into a first sub-group of transmission bits and a second sub-group of transmission bits;
a mapping unit adapted to map the first sub-group of transmission bits onto constellation points of a differential transmit diversity coding scheme for coding the first sub-group of transmission bits;
a coding unit adapted to determine transmission symbols through differential coding of the constellation points and previously transmitted transmission symbols for setup of a transmission symbol vector; and
a scaling unit adapted to scale a length of the transmission symbol vector for coding the second sub-group of transmission bits; wherein
the mapping unit is adapted to map the 2·log 2 (M 1 ) transmission bits onto a constellation vector [A k B k ] of the differential transmit diversity coding scheme according to:
A k =d 2t+1 d (0)*+ d 2t+2 d (0)*
B k =d 2t+1 d (0)+ d 2t+2 d (0)
wherein
d i are constellation elements of a M 1 -ary phase shift keying (PSK) modulation scheme;
d(0) is a freely selectable reference point of the M 1 -ary phase shift keying (PSK) modulation scheme; and
the length of the constellation vector is unit length |A k | 2 +|B k | 2 =1,
the coding unit is adapted to achieve differential coding according to:
( s 2t+1 s 2t+2 )= A k ( x 2t−1 x 2t )+ B k (− x 2t * x 2t−1 * )
wherein
t is an index in time; and
(x 2t−1 x 2t ) is a transmission symbol vector of a previously transmitted matrix; and
the scaling unit comprises a length modification unit, which is adapted to scale the transmission symbol vector length according to:
( x 2t+1 x 2t+2 )=√{square root over (a q k )} ·s 2t+1 s 2t+2 )
wherein
a is a constant; and
q k ∈{−M 2 +1, −M 2 +2, . . . , 0, 1, . . . , M 2 −1} is a length exponent.
12. The apparatus according to claim 11 , wherein the sub-group of transmission bits comprises 2 log 2 (M 1 )+log 2 (M 2 ) bits, wherein
M 1 is the number of possible bit sequences of the first sub-group of transmission bits;
M 2 is the number of possible length values of the transmission symbol vector; and
the dividing unit comprises:
a first selecting unit adapted to select a number of 2 log 2 (M 1 ) bits in the group of transmission bits for the first sub-group of transmission bits; and
a second selecting unit adapted to select a number of log 2 (M 2 ) bits in the group of transmission bits for the second sub-group of transmission bits; wherein
the scaling unit comprises a length exponent memory unit adapted to store at least one length exponent as a function of a bit pattern of the second sub-group of transmission bits through:
calculating a set of M 2 scaling factors δ k ε{0, . . . , M 2 } from a set of bit patterns [p 1 , . . . , P log2(M2) ] covering permutations of the second sub-group of transmission bits for all p 1 ε{0, 1} according to:
δ
k
=
∑
i
=
1
i
=
log
2
(
M
2
)
[
p
i
·
2
i
]
;
and
establishing a pre-determined relation between bit patterns of the second sub-group of input bits [u k,2,l , . . . , u k,2,log 2 (M 2) ] and the set of scaling factors, wherein u k,2,i, is the i-th bit in the second sub-group of transmission bits corresponding-to the constellation vector [A k B k ].
13. The apparatus according to claim 12 , characterized in that the scaling unit comprises a length exponent calculation unit adapted to calculate a length exponent through:
mapping transmission bits of the second sub-group of transmission bits [u k,2,1 , . . . , u k,2,log 2 (M 2 ) ] to a scaling factor δk according to the pre-determined relation established between bit patterns of the second subgroup of transmission bits [u k,2,1 , . . . , u k,2,log 2 (M 2 ) ] and the set of scaling factors; and
calculating the next length exponent according to
q k =δ k −s ( Q k−1 +δ k −M 2 )· M 2 ;
wherein
s( ) is a step function having a value of 1 for non-negative arguments and a value of 0 for negative arguments; and
Q k−1 length exponent representing the absolute length of the transmission symbol vector transmitted prior to calculation of the length exponent q k .
14. An apparatus for differential multiple-length diversity reception of transmission symbols using at least one reception antenna, wherein transmission symbols carry information being coded through mapping of a first sub-group of transmission bits onto constellation points of a differential transmit diversity scheme for coding the first group of transmission bits, subsequent differential coding of the constellation points and previously transmitted transmission symbols for setup of a transmission symbol vector, and scaling of a length of the transmission symbol vector for coding a second sub-group of transmission bits, the apparatus being characterized by:
a vector building unit adapted to organize transmission symbols into a plurality of reception vectors according to a pre-determined scheme;
a combining unit adapted to combine the reception vectors for determination of at least a first decision variable and a second decision variable in relation to the first sub-group of transmission bits and further for determination of a third decision variable in relation to the second sub-group of transmission bits;
a first output detector adapted to determine a first detection output in relation to the first sub-group of transmission bits on the basis of the first decision variable and the second decision variable, respectively; and
a second output detector adapted to determine a second detection output in relation to the second sub-group of transmission bits on the basis of the third decision variable;
wherein
the vector building unit is adapted to organize transmission symbols into a plurality of reception vectors according to:
y
k
=
[
𝓎
2
t
-
1
(
1
)
𝓎
2
t
(
1
)
=
⋮
𝓎
2
t
-
1
(
n
R
)
𝓎
2
t
(
n
R
)
=
]
,
y
k
+
1
=
[
𝓎
2
t
+
1
(
1
)
𝓎
2
t
+
2
(
1
)
=
⋮
𝓎
2
t
+
1
(
n
R
)
𝓎
2
t
+
2
(
n
R
)
=
]
,
y
_
k
=
[
𝓎
2
t
(
1
)
-
𝓎
2
t
-
1
(
1
)
=
⋮
𝓎
2
t
(
n
R
)
-
𝓎
2
t
-
1
(
n
R
)
=
]
,
wherein
t is a time index;
n R is the number of receiption antennas;
* is a complex conjugate operator; and
y i (j) is a symbol received at time i at reception antenna j.
15. The apparatus according to claim 14 , wherein the combining unit is adapted to combine the reception vectors for determination of a first decision variable ŷ 1 , a second decision variable ŷ 2 and a third decision variable ŷ 3 according to:
y
^
1
=
y
k
H
y
k
+
1
;
y
^
2
=
y
k
H
y
k
+
1
;
and
y
^
3
=
y
k
+
1
H
y
k
+
1
y
k
H
y
k
;
wherein
H is operator of transposing a vector and applying the conjugate complex operator * to all vector elements.
16. The apparatus according to claim 14 , wherein the first output detector is adapted to operate in a hard detection mode;
the first output detector comprises a constellation matching unit adapted to determine the first detection output as constellation vector [A(i)B(i)] from the differential transmit diversity scheme which is closest to a vector set up from the first decision variable and the second decision variable [ŷ 1 ŷ 2 ]
according to:
( Â k {circumflex over (B)} k )=arg min i {|ŷ 1 −A ( i )| 2 +|ŷ 2 −B ( i )| 2 }.
17. The apparatus according to claim 16 , further comprising a first bit demapping unit adapted to obtain the first sub-group of transmission bits through demapping from (Â k {circumflex over (B)} k ).
18. The apparatus according to claim 14 , wherein the second output detector is adapted to operate in a hard detection mode; and
the second output detector comprises a scaling factor detection unit adapted to determine a length exponent a qk which is closest to the third decision variable ŷ 3 according to:
â q k =arg min i |ŷ 3 −a q(i) |, q ( i )∈{− M 2 +1, . . . , −1,0,1, M 2 −1}
wherein
{−M 2 +1, . . . , −1, 0, 1, M 2 −1} is a set of all candidate length exponents; and
a is a constant.
19. The method according to claim 14 , wherein the first output detector is adapted to work in a soft detection mode; and
the first output detector comprises a first log likelihood calculation unit adapted to determine log-likelihood ratios for the first sub-group of transmission bits according to
L
(
1
)
(
u
^
k
,
l
)
=
log
p
(
u
k
,
l
=
+
1
|
𝓎
^
1
,
𝓎
^
2
)
p
(
u
k
,
l
=
-
1
|
𝓎
^
1
,
𝓎
^
2
)
wherein
k is a time index;
a vector of the first sub -group of transmission bits u k of dimension 2log 2 (M 1 ) is mapped onto one of the M 1 2 constellation elements of the differential transmit diversity scheme and u k,l is a transmission bit at position l in u k ;
ŷ 1 is the first decision variable;
ŷ 2 is the second decision variable;
p(u k,l =+1|Ŷ 1 , ŷ 2 ) is a conditional probability for u k,l =+1 in view of determined decision variables ŷ 1 and ŷ 2 ;
p(u k,l =−1|ŷ 1 ,ŷ 2 ) is a conditional probability for u k,l =−1 in view of determined decision variables ŷ i , and ŷ 2 ; and
L (1) (u k,l ) is the soft output for the first sub-group of transmission bits.
20. The apparatus according to claim 14 , wherein the second output detector is adapted to operate in a soft detection mode; and
the output detector comprises a second log likelihood ratio calculation unit adapted to determine log-likelihood ratios for the second sub-group of transmission bits according to
L
(
2
)
(
u
^
k
,
l
)
=
log
p
(
u
k
,
l
=
+
1
|
𝓎
^
3
)
p
(
u
k
,
l
=
-
1
|
𝓎
^
3
)
wherein
k is a time index;
u k,l is a transmission bit at position l in a vector u k of dimension log 2 (M 2 ), the vector u k being set up from the second sub-group of transmission bits;
ŷ 3 is the third decision variable;
p(u k,l =+1|ŷ 3 ) is a conditional probability for u k,l =+1 in view of the decision variable ŷ 3 ;
p(u k,l =−1|ŷ 3 ) is a conditional probability for u k,l =−1 in view of the decision variable ŷ 3 ;
and
L 2 (u k,l ) is the soft output for the second sub-group of transmission bits.
21. A computer program stored on a computer readable storage medium wherein the program is directly loadable into an internal memory of a differential multiple length diversity transmitter comprising software code portions for performing the steps of claim 1 , when the program is run on a processor of the differential multiple length diversity transmitter.
22. A computer program stored on a computer readable storage medium wherein the program is directly loadable into an internal memory of a differential multiple length diversity receiver comprising software code portions for performing the steps of claim 4 , when the program is run on a processor of the differential multiple length diversity receiver.Cited by (0)
No later patents cite this yet.
References (0)
No backward citations on record.